Discovery by a proteomic approach of possible early biomarkers of drug-induced nephrotoxicity in medication-overuse headache

Background

Medication-overuse headache (MOH) is a chronic headache condition that results from the overuse of analgesics drugs, triptans, or other antimigraine compounds. The epidemiology of drug-induced disorders suggests that medication overuse could lead to nephrotoxicity, particularly in chronic patients. The aim of this work was to confirm and extend the results obtained from a previous study, in which we analyzed the urinary proteome of 3 MOH patients groups: non-steroidal anti-inflammatory drugs (NSAIDs), triptans and mixtures abusers, in comparison with non-abusers individuals (controls).

Methods

In the present work we employed specialized proteomic techniques, namely two-dimensional gel electrophoresis (2-DE) coupled with mass spectrometry (MS), and the innovative Surface-Enhanced Laser Desorption/Ionization Time-of-Flight mass spectrometry (SELDI-TOF-MS), to discover characteristic proteomic profiles associated with MOH condition.

Results

By 2-DE and MS analysis we identified 21 over-excreted proteins in MOH patients, particularly in NSAIDs abusers, and the majority of these proteins were involved in a variety of renal impairments, as resulted from a literature search. Urine protein profiles generated by SELDI-TOF-MS analysis showed different spectra among groups. Moreover, significantly higher number of total protein spots and protein peaks were detected in NSAIDs and mixtures abusers.

Conclusions

These findings confirm the presence of alterations in proteins excretion in MOH patients. Analysis of urinary proteins by powerful proteomic technologies could lead to the discovery of early candidate biomarkers, that might allow to identify MOH patients prone to develop potential drug overuse-induced nephrotoxicity.

The role of the kidney in lipid metabolism

PURPOSE OF REVIEW

Cellular uptake of plasma lipids is to a large extent mediated by specific membrane-associated proteins that recognize lipid-protein complexes. In the kidney, the apical surface of proximal tubules has a high capacity for receptor-mediated uptake of filtered lipid-binding plasma proteins. We describe the renal receptor system and its role in lipid metabolism in health and disease, and discuss the general effect of the diseased kidney on lipid metabolism.

RECENT FINDINGS

Megalin and cubilin are receptors in the proximal tubules. An accumulating number of lipid-binding and regulating proteins (e.g. albumin, apolipoprotein A-I and leptin) have been identified as ligands, suggesting that their receptors may directly take up lipids in the proximal tubules and indirectly affect plasma and tissue lipid metabolism. Recently, the amnionless protein was shown to be essential for the membrane association and trafficking of cubilin.

SUMMARY

The kidney has a high capacity for uptake of lipid-binding proteins and lipid-regulating hormones via the megalin and cubilin/amnionless protein receptors. Although the glomerular filtration barrier prevents access of the large lipoprotein particles to the proximal tubules, the receptors may be exposed to lipids bound to filtered lipid-binding proteins not associated to lipoprotein particles. Renal filtration and receptor-mediated uptake of lipid-binding and lipid-regulating proteins may therefore influence overall lipid metabolism. The pathological mechanisms causing the pronounced atherosclerosis-promoting effect of uremia may involve impairment of this clearance pathway.

Glomerular protein sieving and implications for renal failure in Fanconi syndrome

BACKGROUND

Glomerular sieving coefficients (GSCs) of proteins have been measured extensively in animals but not humans. We have studied the proteinuria of Fanconi syndrome, a "knock-out" of renal tubular protein reabsorption, to estimate GSCs and detect potential contributors to development of renal failure.

METHODS

Immunoassay of proteins and polypeptides in serum and urine of patients with early Dent's disease (mean GFR = 83 mL/min, range 60 to 101, N = 5), Lowe's syndrome (N = 3), and ADIF (N = 2) were used.

RESULTS

Twenty-one proteins, ranging in mass from insulin (5.1 kD) and parathyroid hormone (PTH; 9.4 kD) to transferrin (78 kD) and intact IgG (160 kD), were present in Fanconi urine at> 6 to 1000-fold normal. A simple model assuming complete "knock-out" of the reuptake of each protein filtered normally by the glomerulus was applied to protein excretion by Dent's patients. GSCs were estimated for 12 plasma proteins, including albumin (7.7 +/- 0.9 x 10-5) and IgG (4.2 +/- 0.28 x 10-5; mean +/- SEM). We calculated the albumin concentration in normal glomerular filtrate to be 3.5 +/- 0.41 mg/L (53 +/- 6.4 nmol/L), consistent with studies in rat and dog.

CONCLUSIONS

To our knowledge, this study provides the first estimates of human in vivo GSCs. Our model explains why tubular proteinuria of Fanconi syndrome includes proteins of mass of albumin and above as well as low-molecular-weight proteins, and further characterizes the endocytic pathway(s) believed defective in these syndromes. High urinary concentrations of potentially bioactive hormones such as PTH, insulin, IGF-1 and the chemokine monocyte chemoattractant protein-1 (MCP-1), were found; their presence in tubular fluid may contribute to the hypercalciuria, interstitial fibrosis, and the progressive renal failure of Fanconi syndromes.

Tubular proteinuria defined by a study of Dent's (CLCN5 mutation) and other tubular diseases

Tubular proteinuria defined by a study of Dent's ( CLCN5 mutation) and other tubular diseases.

BACKGROUND

The term "tubular proteinuria" is often used interchangeably with "low molecular weight proteinuria" (LMWP), although the former implies a definite etiology. A specific quantitative definition of tubular proteinuria is needed, and we address this by studying five different renal disorders.

METHODS

Tubular proteinuria was assessed by measuring urinary retinol-binding protein (RBP), beta2-microglobulin (beta2M), alpha1-microglobulin (alpha1M), and albumin in 138 patients: 26 affected males and 24 female carriers of the X-linked syndrome "Dent's disease," 6 patients with other Fanconi syndromes, 17 with distal renal tubular acidosis (dRTA), 39 with glomerulonephritis (GN), and 26 with Chinese herbs nephropathy (CHN).

RESULTS

RBP was better than beta2M or alpha1M in identifying the tubular proteinuria of Dent's disease. Median urinary RBP levels in mg/mmol creatinine were: affected male Dent's, 18.2, N = 26; carrier female Dent's, 0. 30, N = 24; dRTA, 0.027, N = 17; GN, 0.077, N = 39; and normal adults, 0.0079, N = 61. Elevated urinary RBP (>0.017) and albumin < (10 x RBP) + 2 identified all patients with the LMWP of Dent's disease and clearly distinguished their LMWP from that of dRTA and GN. This is a quantitative definition of tubular proteinuria. Consistent with this definition, 80% of those patients with CHN who had an elevated RBP had tubular proteinuria. Urinary RBP and albumin in carriers of Dent's disease were strikingly correlated over a 100-fold range (R = 0.933).

CONCLUSION

The combination of elevated urinary RBP (>0.017) and albumin < (10 x RBP) + 2 (mg protein/mmol creatinine) is a quantitative definition of tubular proteinuria. Furthermore, our findings suggest that a shared defect in tubular RBP and albumin reuptake causes this form of proteinuria.

Clinical analysis of human urinary proteins using high resolution electrophoretic methods

The application of isoelectric focusing (IEF), sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), two-dimensional electrophoresis (2-DE) and capillary electrophoresis (CE) for high resolution electrophoretic analysis of human urinary proteins is reviewed. In each case, the information is tabulated chronologically with details of sample preparation, electrophoretic system, detection method and clinical application. The text includes an historical perspective of the use of each method for urinalysis and a detailed review of the application of the methods to the investigation of renal disease, renal transplantation, Bence Jones proteinuria (BJP), diabetes mellitus, cadmium toxicity, nephrolithiasis and cancers of the urogenital tract.

Sensitive assays for urinary retinol-binding protein and beta-2-glycoprotein-1 based on commercially available standards

Measurements of urinary retinol-binding protein (RBP) and beta-2-glycoprotein-1 (beta 2G1) provide sensitive indicators of proximal renal tubular malfunction. beta 2G1 is very stable in urine but RBP, which is much less stable, has been more extensively studied. We have designed two similar immunoassays allowing measurement of both proteins from one urine dilution. The assays both use widely available commercial standards and permit accurate measurement throughout most of the reference range. We measured both proteins in the urine of 70 healthy adults and 80 healthy children aged between three and 16 years, expressing the results as a ratio to creatinine concentration. Using the 98th centile as the upper reference limit, adults excreted up to 17 micrograms/mmol RBP (median 8) and up to 30 micrograms/mmol beta 2G1 (median 15) with no difference between the sexes. Children had different ranges to adults with excretion of up to 22 micrograms/mmol RBP (median 10) and up to 80 micrograms/mmol beta 2G1 (median 24).

Identification of beta2-glycoprotein I as a membrane-associated protein in kidney: purification by calmodulin affinity chromatography

Outer renal medulla calmodulin-binding proteins from a soluble protein fraction and a plasma membrane fraction solubilized in CHAPS were retained on a calmodulin-Sepharose 4B column in the presence of Ca2+, and subsequently eluted by EGTA. The calmodulin-binding proteins constituted 2.5% of the soluble protein and 0.1% of the solubilized membrane protein. beta2-glycoprotein I was identified as a calmodulin-binding protein both by N-terminal sequencing and by immunoblotting. Quantification showed that beta2-glycoprotein I constituted the major part (approx. 35%) of the calmodulin-binding membrane proteins, but only a minor part (approx. 0.1%) of the calmodulin-binding proteins in the soluble fraction. These results show for the first time that beta2-glycoprotein I binds calmodulin and that beta2-glycoprotein I may in kidney be a membrane-associated protein. Immunohistochemical studies identified beta2-glycoprotein I in several parts of the cortex and the medulla of the kidney, including Bowman's capsula, the tubular lumen and the tubular epithelium, indicating that beta2-glycoprotein I, despite its relatively high molecular mass, is filtrated in the glomerulus and subsequently reabsorbed by the tubular epithelium. This is in agreement with beta2-glycoprotein I being a marker for renal tubular disease.

Apolipoprotein H levels in diabetic subjects: correlation with cholesterol levels

To assess the relationship between apolipoprotein H (apo H) plasma levels and lipid metabolism in diabetes mellitus, we have examined the correlation between apo H plasma concentration and the main plasma lipid levels in 127 non-insulin-dependent (NIDDM) and 118 insulin-dependent (IDDM) diabetes mellitus patients. The data are compared with those in 286 nondiabetics. Our data show a significant increase in plasma apo H in diabetic as opposed to nondiabetic subjects (NIDDM, 29.9 +/- 10.8 mg/dL; IDDM, 31.3 +/- 9.9; controls, 22.5 +/- 7.7; F = 53.3, P = .0001). The relation between plasma lipids and apo H was simultaneously evaluated in the three groups with inclusion of diabetes, sex, body mass index (BMI), and age as covariates in the model. This analysis showed a strong positive correlation (P = .0009) between apo H and total cholesterol, and a weaker positive correlation with triglycerides ([TGs] P = .016). The correlation between apo H and hemoglobin A1c (HbA1c) levels in diabetics (P = .03) highlights the importance of glycemic control for plasma levels of this apoprotein, which is highly glycated. Although the role of apo H in lipid metabolism is still uncertain, recent investigations on the possible relation between plasma apo H levels and increased plasma lipids and thrombotic risk could explain the increased atherosclerotic risk in diabetic patients.

Immunoelectrophoresis and ELISA techniques for assay of plasma beta 2 glycoprotein-1 and the influence of plasma lipids

Beta 2 glycoprotein-1 (beta 2GP1) has been identified as a cofactor for the binding of some antiphospholipid antibodies to anionic phospholipids and has been demonstrated to possess anticoagulant properties in vitro. We have investigated Laurell rocket immunoelectrophoresis (IEP) and ELISA techniques for measurement of beta 2GP1. Western blotting and crossed immunoelectrophoresis (CIE) of plasma demonstrated free beta 2GP1 and beta 2GP1 complexed with unidentified plasma constituents. The free and complexed forms were not distinguished in immunoelectrophoresis assays, allowing measurement of total beta 2GP1. Standard and detergent modified IEP and ELISA techniques were compared: significant correlation was demonstrated between unmodified and detergent modified IEP, detergent modified IEP and ELISA and unmodified IEP and ELISA. The intra-assay co-efficients of variation (CVs) of the unmodified and modified IEPs and ELISA were 7.5, 3.7 and 7.9%. Inter-assay CVs determined for the modified IEP and ELISA were 5.8 and 9.1% respectively. The purified beta 2GP1 used to standardise the assays was shown to have subfraction selectivity and different calibration values were obtained for pooled normal plasma by the unmodified IEP (242 mg/l) and modified IEP and ELISA (201 mg/l). We have also investigated the influence of some pre-test variables on beta 2GP1 levels and shown that heparin, citrate or EDTA plasma and serum samples are suitable for assay of this glycoprotein and that levels are unaffected by repeated freeze/thawing. The influence of plasma lipids on beta 2GP1 measurement was also examined and we demonstrated no significant differences between pre and postprandial samples, which suggests that fasting status is not an important consideration for assay of beta 2GP1 in healthy subjects.

Beta 2-glycoprotein-1 (apolipoprotein H) excretion and renal tubular malfunction in diabetic patients without clinical proteinuria

AIM

To compare the urinary excretion of beta 2-glycoprotein-1 with that of two other markers of early tubular disorder in diabetic patients without clinical proteinuria.

METHODS

The urinary excretion of retinol binding protein, beta 2-glycoprotein-1, and N-acetyl-beta-D-glucosaminidase was measured in 90 known diabetic patients who had a negative reagent strip test for proteinuria.

RESULTS

Among 43 patients with urinary albumin excretion within the reference range, 23 (53%) had raised urinary N-acetyl-beta-D-glucosaminidase activity, five (12%) increased excretion of beta 2-glycoprotein-1, and five (12%) increased loss of retinol binding protein. Among 47 patients with an albumin excretion of 0.9-7.9 mg/mmol creatinine, 42 (89%) had increased urinary N-acetyl-beta-D-glucosaminidase, 23 (49%) an increased output of beta 2-glycoprotein-1, and 16 (34%) a raised excretion of retinol binding protein. The excretion of these markers of tubular defects seldom exceeded two and a half times the upper reference limit and the differences between the findings in the insulin dependent and non-insulin dependent patients with similar albumin excretion were small and insignificant.

CONCLUSIONS

In diabetic patients with a negative dipstick test for proteinuria: (a) assay of urinary beta 2-glycoprotein-1 may be a more sensitive test for the detection of impaired tubular reabsorption of protein than measurement of retinol-binding protein; (b) assay of N-acetyl-beta-D-glucosaminidase can detect tubular injury at a time when protein reabsorption remains normal; and (c) impaired renal tubular function may be present in the absence of evidence of glomerular malfunction.

Urinary excretion of beta 2-glycoprotein-1 (apolipoprotein H) and other markers of tubular malfunction in "non-tubular" renal disease

AIM

To determine whether urinary beta 2-glycoprotein-1 assays can provide improved discrimination between chronic renal diseases which are primarily of tubular or glomerular origin.

METHODS

Urinary beta 2-glycoprotein-1, retinol-binding protein, alpha 1-microglobulin, beta 2-microglobulin, N-acetyl-beta-D-glucosa-minidase and albumin were measured in 51 patients with primary glomerular disease, 23 with obstructive nephropathy, and 15 with polycystic kidney disease, and expressed per mmol of creatinine. Plasma beta 2-glycoprotein-1 was assayed in 52 patients and plasma creatinine in all 89. The findings were compared between the diagnostic groups and with previously published data relating to primary tubular disorders.

RESULTS

All 31 patients with plasma creatinine greater than 200 mumol/l excreted increased amounts of beta 2-glycoprotein-1, retinol-binding protein, and alpha 1-microglobulin, and 29 had increased N-acetyl-beta-D-glucosaminidase; the quantities were generally similar to those found in comparable patients with primary tubular pathology. Among 58 with plasma creatinine concentrations under 200 mumol/l, increases in beta 2-glycoprotein-1, retinol-binding protein, and alpha 1-microglobulin excretion were less common and much smaller, especially in those with obstructive nephropathy and polycystic disease. The ratios of the excretion of albumin to the other proteins provided the clearest discrimination between the patients with glomerular or tubular malfunction, but an area of overlap was present which embraced those with obstructive nephropathy and polycystic disease.

CONCLUSIONS

Increased excretion of beta 2-glycoprotein-1 due to a raised plasma concentration or diminution of tubular reabsorption, or both, is common in all the forms of renal disease investigated, and both plasma creatinine and urinary albumin must be taken into account when interpreting results. Ratios of urinary albumin: beta 2-glycoprotein-1 greater than 1000 are highly suggestive of primary glomerular disease and those less than 40 of primary tubular disease. Used in this way, beta 2-glycoprotein-1 assays provide superior discrimination between glomerular and tubular malfunction when compared with retinol binding protein but the best discrimination is provided by albumin: alpha 1-microglobulin ratios.

Beta 2-glycoprotein-1 (apolipoprotein H) excretion in chronic renal tubular disorders: comparison with other protein markers of tubular malfunction

Urinary beta 2-glycoprotein-1 was measured in 60 patients with conditions recognised as causing renal tubular impairment and compared with established markers of early tubular malfunction. Increased beta 2-glycoprotein-1 excretion was found in 49 (82%) of the subjects; raised excretion of alpha 1-microglobulin, retinol-binding protein, and beta 2-microglobulin was found in 46 (77%), 45 (75%), and 31 (52%), respectively, and increased urinary N-acetyl-beta-D-glucosaminidase activity in 32 of 54 of the subjects (59%). The increase was particularly pronounced in those with proximal tubule malfunction, although considerable variation occurred. beta 2-glycoprotein-1 was shown to be stable in urine over the physiological pH range, and it is concluded that its measurement provides a means of detecting chronic malfunction of the renal tubules that is marginally more sensitive than assays of alpha 1-microglobulin or retinol-binding protein, and more reliable than assays of beta 2-microglobulin or N-acetyl-beta-D-glucosaminidase.

A sandwich enzyme-linked immunosorbent assay for beta 2-glycoprotein I

A solid-phase sandwich enzyme-linked immunosorbent assay for determining beta 2-glycoprotein I in urine has been developed. It has a working concentration range of 5-40 micrograms/L and a detection limit of approximately 1.4 micrograms/L. The within-plate coefficient of variation (CV) falls between 1.4% and 2.1%, and the between-batch CV ranges from 5.2 to 6.0%. Recovery of beta 2-glycoprotein I added to urine varies between 96 and 110%. The assay can also be used for determining beta 2-glycoprotein I in serum.